Testing mosfet, won't fully discharge?

[xboxgamer]

So please help me understand why my mosfet does not act as expected.

I am testing mosfet out of circuit, not installed yet.

I switch multimeter to diode mode. I put negative probe on source and positive on gate. The voltage drop is as expected 0.000 between drain and source.

Now is the part I don't understand. All articles or youtube videos mention that in order to turn the mosfet off all I need to do is to short gate and source by either finger or metal.

No matter if I use finger or pliers or whatever I don't see OL between drain and source. I see random readings like 2.6V to 2.9V voltage drop.

However (I came to this by trial and error testing random stuff) if I put positive probe of my multimeter on drain and negative on gate, the mosfet turns off completely immediately and drain and source shows OL.

Please help me understand why finger shorting method does not show OL and shows high voltage drop vs when I put negative probe on gate and positive on drain it turn off as expected.

Thanks 

[DavidAlfa]

Discharge it by inverting the multimeter tips, positive to drain, negative to gate, then it'll be completely off.
It heavily depends on the MOSFET, some are really sensitive.

Ground yourself, MOSFET gates can get damaged quite easily, anything over 20V is risky.
The gate capacitance is so small that it can pick dangerous voltage levels just by touching it due capacitive coupling.
I remember wasting few 2n7002 this way, worked until touched with my fingers, then it was gone, permanently shorted.
After damaging a few more, I set the DMM to measure ac voltage between myself and the socket's earth, was something like 140volts!

It's only a few micro/nanoamps, you don't feel anything, but enough to charge and kill a small MOSFET.